The ?Computational Models of Immunity? projects in this application focus on development and implementation of new structure-based design tools for influenza hemagglutinin (HA) protein trimer interface specific antibodies or vaccine antigens. These projects will use knowledge about the structure and function of human neutralizing antibodies to the trimer interface of the HA head that we have in hand or will discover, in order to design new antibodies or vaccines in silico. We have access to peripheral blood cells from a diverse panels of subjects with prior natural infection, or exposure to experimental inoculation with vaccines encoding HA molecules with both seasonal vaccines and unusual experimental influenza subtypes, including H3variant, H5, H6, H7, H9, and H10 viruses. The immune B memory cell populations from these individuals are the ideal starting materials to isolate unusual heterosubtypic antibodies. Recently, we identified the HA head trimer interface as a major new site of vulnerability for universal influenza antibodies and candidate vaccines. Here, we will study existing and isolate additional broadly heterosubtypic human antibodies to the trimer interface of the HA head. We will determine the immunome of the responding heterosubtypic clones using high-throughput next generation sequencing of antibody gene repertoires that comprise the clonal lineages of the most heterosubtypic antibodies isolated. Once antibodies with unusual breadth or activity are isolated, the structure of these antibodies will be determined in complex with purified HA molecules in the Structural Core using crystallography and single particle electron microscopy (EM) studies. Such structures will provide the coordinates for the modeling experiments using Rosetta. We will in silico mature human antibodies to increase affinity for the HA antigen of specific virus types and use multi-state design to maximize breadth, i.e., create antibodies that recognize HAs of all clades, subtypes, groups, or even types. We then will synthesize and express these novel antibodies and determine neutralization activity, binding affinity, and competition binding groups of designed antibodies, using a diverse HA panel and pseudotyped viruses with all type A HAs in nature. The co-crystal structure of these human antibodies with HA will be the template for in silico design of structurally stable epitope-focused immunogens. We will first validate these designed immunogens by testing the interaction with the target human antibodies. Further, these immunogens will be experimentally tested by evaluating immune responses. Then, we will use the novel immunogens to isolate new antibodies from subjects naturally exposed to influenza, to show that the immunogens present antigens recognized by natural immune responses.